_

*,

'c ..*T .GERRY E. STUDDS. ,' WCdiHINGTON* ' , f,,. TtMTu Disfatt. M AssAcwussive * *g , , g 3

' 202-225-4I11' COMMITTfE ON MERCHANT MARINE - * SOtJTH GHORE 1-800 4 94-9911,' AND FISH (HtES

af ,, ,2 =: .....,'--se-o . m o, E _. .M, , e- o- MA - ,..

' AND NArunAt Renovacts .anocarONCH A69 MAN FEDEhAL Burtowso

Congregg of tfje Eniteb fptates = = >Co-mE ON EN,a-AND COMMERCE ptvuoVTH

S"|||";"',':,;" "'';'" Riottgr of Representatibts '"C''L'Cl ', ,,SUSCOMMifTit ON TRANSPORT ATs04 h

9AND HAJA#Dous MAT &PtALs

Hvament. MA 02601

March 1, 1994

Dennis RathbunDirector, Office of Congressional Affairs'Nuclear Regulatory CommissionWashington

,

DC 20555

I am contacting you on behalf of my constituent, Ms. Mary Ott.Please see enclosed correspondence.

I would appreciate your attention to Ms. Ott's concerns,specifically the discrepancies between the attached memo and yourresponse to my earlier inquiry on this matter. Please respond tome at: 1212 Hancock St, Quincy, MA 02169, Attm Mary Lou Butler.

'

S i.nce J y,N

3

oGe E St dds.

Enclosure

I-l

;

|

|-i.

*

|

9405050328 940418PDR ADOCK 05000293 ,PDR ,p

,1 ,'

THIS STATIONERY PRINTED ON PAPER MADE OF RECYCLED FIBERS

k'

,

_.

.;.- .:. .

,,.

Citizens Urging Responsible Energy-_

Jc -

February 17, 1994

The Honorable Gerry E. StuddsUnited States House of RepresentativesWashington, DC 20515-2110

Dear Congressman Studds:

In his February 1, 1994 letter to you on behalf of the Union ofConcerned Scientists (UCS), nuclear safety expert Robert Pollard citescompelling new evidence that raises questions about the justificationfor continued operation of the Pilgrim Nuclear Power Station (PIIGRIM)in Plymouth.

UCS cautions that a turbine failure at PIIGRIM is not merely an economicconcern for the Boston Edison Company, as the Nuclear RegulatoryCommission (NRC) suggests, but "could result in a serious nuclearaccident with significant radiological consequences for the public."These are sobering considerations for your constituents in Massachusettsand all New England as well.

We are writing to support UCS's request for additional analysis of therisks posed by PIIGRIM's cracked turbine, to provide information >refuting claims by Boston Edison, and to request an investigation todetermine whether the NRC attempted to deceive you about the risks posedby continued operation of the PIIGRIM plant.

andThe Christmas day turbine fa:,1ure at the Fermi 2 nuclear plant,iluresrecently issued fGC Informat: on Notice 64-01, " Turbine Blade FaCaused by Torsional Excitation (i.e., vibration) From Electrical SystemDisturbance," offer insights into UCS's concerns with the potentialdangers of PIIGRIM's cracked main turbine.

Boston Edison dismisses the lessons to be learned from the turbinefailure at the Fermi plant by stating that the Femi turbine'was made by -a different manufacturer than the PIIGRIM turbine. However, UCS citedthe Fermi turbine failure as evidence that a failure of an_y,

~ |turbine can cause large vibrations. In the case of PI1 GRIM, thevibrations could increase the force on the existing cracks, but the NRC

!analysis of PIIGRIM did not consider this.

Similarly, UCS cited NRC Information Notice 94-01,which describes howroutine electrical disturbances can cause turbine vibrations that:wouldsignificantly increase the force on the cracks in PIIGRIM's turbine.

;

;

(617) 934-0498'Duxbury, Massachusetts 02331 .Post office _ Box 2621 -

_ _

.'... , ' . , ' ,'

. .

.

.

Boston Edison has denied that PI1 GRIM is subject to the electricalHowever, our research of NRCdisturbances the NRC varned against. By 1987, the NRC had documentedreports contradicts Edison's claim. In the20 events which caused loss of 345kV offsite power at PIIERIM.

Eeh"7 M6 kSe specih[$aESfc[Ech b"hieECNNe NpNo 2

as having the potential to perturb the existing cracks.We

Finally, Congressman Studds, somothing else is seriously wrong here.17, 1993 NRC internal memo, "PIIGRIM UNIT

have obtained a copy of a JuneASSESSMENT OF IDW PRESSURE TURBINE ANALYSIS," prepared for ProjectDirector, Walter Butler by Engineering Chief, Jack Strosnider, which1:

was the basis for Chaiman Selin's August 4,1993 response to you andA comparison of the internal memo and.

CURE, " Enclosure 2" (appended).the NRC response to you shows that the NRC is saying one thing on the"inside", and another on the "outside", about the danger of PIIGRIM'scracked turbines.

The NRC's response to you is essentially a verbatim repetition of theinternal memo, but there are glaring changes and omissions which,'atbest, could be viewed as an attempt by the NRC to downplay the safetyAt vorst, they could represent ahazards of PIIGRIM's cracked turbine.deliberate attempt to withhold vital safety information to deceive ourCongressional leaders about potentini dangers of PI1 GRIM's continuedoperation.

The critical importance of turbine deck orientation is outlined in NRCREGUIATORY GUIDE 1 115, Protection Against low-Trajectory Turbine

It warns, "..it is necessary toMissiles, provided to you by the UCS. Itshow that the risk fran turbine missiles is acceptably small."recites five principal means of safeguarding against such missiles andconcludes, "...The first of these, turbine orientation and_ placement,provides a high dogree of confidence that low-trajectory missilesresulting from turbine failures will not damage essentihl systems."(e.g., spent fuel storage pool, or impairment of vital' control roomfunctions) emphasis added.

Yet in response to you, the NRC removed key language from the assessmentblywhich reveals that PIIGRIM's "... turbine dock is orientated unfavoraAlso, in its response to you,

with respect to the reactor building."the NRC says, "...(it) desires that the turbine dis}, failure probabilityk

be 1E-5 each year or lower for an unfavorably orientated turbine.But on the "inside", the NRC says, ". . .(it) requires .(emphasis added)that the turbine missile failure probability be 1E-5 per year or lowerfor an unfavorably orientated turbine." (emphasis added)

;

i

_ -

__'

..

./ !.*

i, ,

.

.

On behalf of the 1,600 members of CURE, we ask you to lend the full- support of your office to:

* ceek an immediate cornitment from the NRC to provide allinformation and analyses requested by UCS

* request the Government Accounting Office, the NRC Inspector Generalor other appropriate agency conduct an investication to determine thereasons why the NRC's 6/17793 internal memo differs from the NRC's8/4/93 response to you

In recent weeks, the media has extensively exposed the past abuses ofthe atomic era. There is little comfort in this progress, whencontemporary abuses are flagrant. We believe that the NRC has abdicatedits role as a regulator, abused its authority and violated our trust.{IIERIM'.s continued gargng an unacceptable rggoing " cost versus safetyr tion has become an ottle," and we are sk

These are strong sentiments from a conservative safe energy group. But

cngrs!$nafh6daEhNrYan s"eis$nNakfSes$Neiuni 1Eatomic abuses of today be revisited 50 years from now, and it willbe too late.. Help us be a voice for change.

We greatly appreciate your involvement in our behalf on PILGRIM safetyissues and look forward to hearing from you at your earliestconvenience.

Sincerely yours,

/ d4A& h -.s.Dr. Donald M. Muirhead Jr. Mary C. OttCo-Chairmen

||4

-__m . _ _ _ -

. ., . . . - . . . .. - - -. . ,_ ./*- - - ~

.,

UNino sTAMS, ' 't4* e

k NUCLEAR REGULATORY COMMi$8 ION'.

.+ S w aneu m ore n, o s. se a w o m

/ JUN 17'1993k[,,/

Walter R. autler, Project Director |NEhCSANDUM FORT Ject Directorate I-3 |Pro:Divanion of Reacter Projmote I/II

Jeok R. Stroenider, chief |Meterials and Chemical Engineering BranchFROMt

!Division 4f Engineerinq u!

,

PILGRIN UNIT la Assammeumy op Low :SUBJECT: PRESSURE TURNINE ANAL 1f8IS

General $1ectrio - (GB) /During the refuellag outage in April 1942, ( at Pilgriainspected the rotor in the low pressure turbine , 5, and 7 |

Unit 1 and found flaw indications in disk numbere 4,GE recommended that the licensee either reeeve the 7stage disk

-

!'

beformor w rrm the LPA Storsubsequently, the licenses retained W1 Integrityto evaluate flaw indicationd iin the 7GA

(7GA)on the generator sidestartup.Associates, Inc. (sIA)

1993, the licensee provided to the NRC M oct managerThe project manager requested thatdisk.

on May 12,wteether he, perform athe SIA analysis (aeference 1) . (ENCW1the Materials and chemical Engineering sceneh re is . any 1

review of the SIA analysis to determine 1)the potentiali impact togrose error in the SIA entlysis and 2) |i

d LPR, with fplant safety.Pilgrim Unit i has two low pressure turbines, LPA an-withThe_ turbine dock js_o_rientated

"*P'- ? '"bo 7GA?

,

t

2me s N oations)of tshrunk-on disks. respect to the reactor W idand web. Aithenegh thf|4th and stb 4

.

|h t in tlpe 74& disk,disk are located in teeth the

,

stage disks have more and-larger flaws than t aas determined that the 7GA disk is the limiting disk hamefractura toeghness of the disk, mit potentiald on |

l

operating conditions, disk failure.consequences of the ffects bf fracture4SIA performed parametrio studies to determ na efresturd

toughnesetoeperature (FART), e intensity

transitionming, orsak growth rate end both the SSappearenosvariability, ENCS staff compared key parameters used | Parametersy. m ,

and SIA analyses to our estimates (see Atte-esment 1) EXAp _of- this_Theanalysis Imodels. 'the

used in the 65 analysis were extracted 1senavailable_ at _ tjhebecause GE's analysis wags

- I- -

--

fm eesoment. . I

'

t

| !John Tsaa, on/ENCS, !

Contact:' :so4-2702.

Q(p 0hb.....,...u......j

!

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ __ _ _ _ _ _ _ _ . _ _ _ _ _. --

. . . _ ._ _ _ _ _ _ _ _ _ _

'

.. . .

Jane I'l - /\/RQ, G\CR1Q !''' '~.

2

ed one indication of 0.14 iFor the 7aA disk, os r the web which as could natl nch in theaccurately khub and an indicationmise. For that indication, GE assumed a'orack sizebased on flaw indications from other poser plants' ins (of 0.25 inchpection dataand labornt6ry ' data. The staff hell -that--tha._jaitial orpoksize of 0.25 inch is conservative staff could p'ot quapti 1Wthe uncertainty associated with the sise at th.is time. /p 7-

( < *> ''

4R used a fracture mechanios model of an edge orack in an infinite ,plate having eenstant loading. ea's model is ::z w4ive becauseit la more compliant than the actual geometry which is a radial-crack emanating from the keyway.

Moreover,inits sonatant loadingdoes not consider the radial decrease hoop 6 tress withincreasing distance tros the here.- sIA'a modai la b hele in aninfinite plate with attenuated leading along the ersch, The staff

,j assuesd a model of a thick wall cylinder with attanud$ed loading.

L)M 4E used 0.08 inch par year for the orsak growth rate W iah was thes' median valge-taken from a statistioni study of the average eraskF

growth ratpd vs.p1 operating temperature from turbino laspectiondata of noth _wa and PWR

r.. _

..' ,

..,

*.

..

3 i

The Wac requirse that the turbine missile fmilure probability be1E-5 per ymer or' lower for an unfavorably orientated Nrbine. GB's-analysis is based on a turbine missile generation probability ofit-6 failure per year. 814 | die m* r-- * a= a nrah-hill'stio *fruture mechanics analysis. Scaed os engineering judgement, theEtwrr 1Welsated that una miselle failure probability for the LFAturi, tem 2.s betws4A 18-S and 15-4 per year. For this moition, theNRC pereits the turbine to be kept in servios il the nextschetuled outage, at which time the licensee should &ke action to

~

reduce _the__ failure probability to the 1E-5 per year ctiterion (Ref.

replacement of both LPA and LFB rotore during the n$14ering2).'nia licenses has indicated that they are co the

dxt scheduledrefuelingoutage,whichisexpectedtobeApril199p

3concludse that there is no safety concern for norme(k the statfI ased on an anwessment ofB the informationgilab !9peration of fthe LPA turnine to the end of the currenQuel gy6le, The SIA

. analysis appears to reduce some of the obvioue ~d66,servatisms used,in the Gr analysis. J

~

Jack R.-Stroenider, Chief .I .I Materiale and Chemical Engineering Branch

Division of Engineering |'

Enclosure:Attachment 1: Pilgria Turbine Evaluation !Attechsent 2 Beforencee I

cc: B. D. Liaw | )'R. Fa *:enJ. Shting' ,

i

t

.

#

.

!__ _ _ __

_ _ _ __ _ _ _ . ___. ._ _ ___ _ _ ______ __ ._____________-___________ ____ _ __-_____ - _

. PoO8,1 - -' ' * :,

.

Enclosure 2

PILGRIM UNIT 1: ASSESSMENT OF LOW

PRESSURE TURBINE ANALYSIS

During the refueling outage in April 1993, General Electric (GE) inspected therotor in low pressure turbine "A" (LPA) at Pilgrim Unit I and found flawindications in disks 4, 5, 6, and 7. GE recommended that the licensee eitherremove the seventh stage disk on the generator side (disk 7GA) or warm the LPArotor before starting the turbine. The licensee later retained StructuralIntegrity Associates, Inc. (SIA) to evaluate flaw indications in disk 7GA.

On May 12, 1993, the licensee submitted the SIA analysis (Reference 1) to theNRC project manager, who requested that the NRC Materials and ChemicalEngineering Branch (EMCB) review the SIA analysis to determine: whether therewere any gross error in the SIA analysis and whether the flaws indications inthe turbine disks would have any effect on plant safety.

Pilgrim Unit I has two low pressure turbines, LPA and LPB, with shrunk-ondisks. The flaw indications of the 7GA disk are located in both the hub andweb. Although the fourth and fifth stage disks have more and larger flawsthan the 7GA disk has, GE determined that the 7GA disk is the limiting disk 1based on operating conditions, the fracture toughness of the disk, and.theconsequences of a disk failure.

._

SIA performed parametric studies to determine effects of the fractureappearance transition temperature (FATT), fracture toughness variability, pre-warming, crack growth rate, and stress intensity factors. The EMCB staffcompared key parameters used in both the GE and SIA analyses to our estimates(see Attachment 1). Parameters used in the.GE analysis were extracted fromthe SIA analysis because GE's analysis was not available at the time of thisassessment.

.

For the 7GA disk, GE reported one indication of 3.556 mm (0.14 in) in the huband an indication in the web which GE could not accurately size. For thatindication, GE assumed a crack size of 6.35 mm [0.25 in) based on flawindications from other power plants' inspection data and laboratory data. Thestaff believes that the initial crack size of 6.35 mm (0.25 in) isconservative but could not quantify the uncertainty associated with theassumed size.

.

GE used a fracture mechanics model of an edge crack in an infinite platehaving constant loading. GE's model is conservative because it is morecompliant than the actual geometry, which is a radial crack emanating from thekeyway.

- ---

-

h . ,t ),. ..

,..

... P.09. .;

..- i

-2-

Moreover, its constant loading does not consider the radial decrease in hstress with increasing distance from the bore. oopinfinite plate with attenuated loading along the crackSIA's model is a hole in anmodel of a thick wall cylinder with attenuated loading.The staff assumed a.

GE used a crack growth rate of 1.52 mmmedian value from a statistical study co(0.06] inch each year, which was the

rrelating the average crack growthrate with the wheel operating temperature from turbine inspection data of bothBWR and PWR plants.0.06 in) each year in its studies.SIA used 0.416 mm [0.0164 in), 0.51 mm0.02 in), and1.52 mm

growth ra(te of 0.51 mm (0.02 inch] each year from previous inspection data ofi

The staff calcu' ated a crack |the LPA rotor.The staff believes that the actuhl crack growth rate may bebetween 0.51mm (0.02 in] and 1.52 mm (0.06 in) each year

indicate the upper bound growth rate (2 standard deviations) at an oHowever, GE's data.temperature of 78 *C [172 *F] could be as high as 2 03 mm (0 08 in) peratingeach year..The critical stress intensity (K ) is an indicator of fracture toughnes

.

the disk material. The lower the K s ofie

themoreconservativetheresultswblbe.used in the fracture mechanics analysis

intensity vs. excess temperature (test temperature - FATT)115 HPalm (105 ksi/in) which was taken from the graph of critical stresGE used a lower bound value ofs !that the value of 115 HPalm (105 ksilin) is conservativeThe staff finds

.

.

GE and SIA calculated the critical crack sizes (depths) of 8 64 m (0 34 i )and 13.72 mm (0.54 in), respectively. n.crack length is the length of the keyway bore.SIA conservatively assumed that the

.

crack aspect ratio is known, the critical crack size may be larger thanSIA indicated that if the13.72 mm (0.54 in).about 11.43 mm (0.45 in] for the thick wall cylinder model.SIA's calculation results in a critical crack size of

Using the above parameters, the staff estimated a factor of safety for flawsize ranging from 1.21 to 3.6 based on the ratio between the crack length at!end of the current fuel cycle in April 1995 to that of the critical crack sizeof the cylinder model (see attachment). The factor of safety for stressroot of the safety factor for flaw size. intensity (K ) ranges from 1.1 to 1.89, which was estimated by taking square

g

The NRC desires that the turbine disk failure probability be IE-5 each year or'\lower for an unfavorably orientated turbine. GE's analysiturbine disk failure probability of IE-5 failure per year.s is based on aperform a probabilistic fracture mechanics analysis. SIA did not

judgment, the staff estimated that the turbine disk failure probability forUsing engineeringthe LPA turbine is between IE-5 and IE-4 per year.turbine in this condition to remain in service until the next scheduledThe NRC would permit aoutage, at which time the licensee should ensure they meet the turbine diskfailure probability to the 1E-5 per year criterion (Attachment 2, Ref. 2).

.. ,

7' ' I ''. :' e.to._ . .s

,"

t

-3-

Upon assessing tho information available, the staff found no safety concernfor normal operation of the LPA turbine to the end of the current fuel cyclalthough the SIA analysis is less conservative than the GE analysis.e,staff intends to perform a confirmatory review of the GE analysis and itsThemethodology.

low pressure turbines during the next refueling outage, which is expected toThe Boston Edison Company has informed the 11RC that it will be replacing bothbe in April 1995.

,

&

l

.

i!